Target Discovery Platform Identifies Cardioprotective Genes Using Hibernation Omics Data

Abstract

Over the course of their evolution, many mammals have naturally solved a number of human biomedical problems. The genes underlying their protective strategies are likely neither novel nor species‐specific, but rather shared among all mammals, including humans. Identifying and validating these genes could therefore lead to new therapeutic treatments for human clinical indications. To this end, we developed a target discovery platform which mines multi‐omics data from mammals resistant to human pathologies in order to identify novel therapeutic genetic targets. More specifically, our platform is built upon transcriptomic and epigenomic data from bio‐banked 13‐lined ground squirrel (13‐LGS) tissues collected at precise seasonal and physiological time‐points across their annual cycle of hibernation. Our platform also incorporates myriad genomic, transcriptomic, and proteomic datasets publicly available for this and other disease‐resistant species, as well as human genomics and transcriptomics data. Here we sought to identify novel genetic targets protective against cardiac fibrosis. While the hearts of 13‐LGS experience repeated bouts of ischemia‐reperfusion during hibernation, long‐term fibrotic scarring resulting from this type of injury is not detected. To identify protective targets, we first generated RNASeq data from hearts of 13‐LGS across 11 different seasonal and physiological groups (n≥5 per group). After obtaining transcript abundance counts for each sample library, we performed weighted correlation gene network analysis (WGCNA) to identify modules of transcripts co‐expressed among the different physiological groups. One resulting WGCNA module contained TGFB1 (transforming growth factor beta 1), a known central mediator of fibrogenesis. Knockdown (KD) of this module’s hub gene, designated here as FAUN211G, as well as the second top‐ranking gene, designated FAUNA212G, significantly decreased collagen deposition in an in vitro 2D model of cardiac fibrosis using primary human cardiac fibroblasts. We next perturbed these genes in an in vitro 3D spheroid model of human cardiac fibrosis and detected improved parameters of cardiac function. Finally, we tested small molecule inhibitors on FAUNA211G in both 2D and 3D in vitro models; we identified significant reductions in myofibroblast activation, proliferation and extracellular matrix deposition. Both FAUNA211G and FAUNA212G were previously unknown to be involved in cardiac fibrosis. In sum, we demonstrate that network analysis of the hibernator’s transcriptome is a strategy to identify novel target genes for treatment of human clinical indications.